MCP Agent Server
About
An innovative Model Context Protocol hub that seamlessly connects your client with intelligent agents, enabling powerful cross-platform communication across multiple MCP ecosystems
Explore
- Multiple Agent Support: Create and manage different specialized agents
- Agent Composition: Combine specialized agents into master agents
- Custom Tool Integration: Create your own tools or use existing MCP servers
- Preconfigured Servers: Easy access to popular MCP servers like Sequential Thinking, Brave Search, and Memory
- AI SDK Integration: Support for 17+ LLM providers including OpenAI, Anthropic, Google (Generative AI & Vertex), Amazon Bedrock, Azure, Cohere, Mistral, Fireworks, Groq, Perplexity, Together AI, xAI, DeepSeek, Cerebras, DeepInfra, and Replicate through Vercel AI SDK v5
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
MCP Agent ServerCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Prerequisites
- Node.js (v16+) installed
- An API key for your chosen AI model provider (OpenAI, Anthropic, Google, etc.)
- Any API keys required by specific MCP servers you want to use
1. Prebuilt Servers (recommended):
{
"prebuilt": "sequentialThinking"
}
Available prebuilt servers: sequentialThinking, memory, braveSearch, firecrawlMcp, fetch, awsKbRetrieval, everart, fileSystem, sqlite
2. Custom MCP Servers:
{
"mcpServers": {
"serverName": {
"command": "node",
"args": ["/path/to/server.js"],
"env": {
"API_KEY": "your-api-key"
}
}
}
}
3. Agent References (for master agents):
{
"agentRef": "Other Agent Name"
}
4. Imported Agents (for using external agent files):
{
"type": "import",
"importPath": "/path/to/your/agent-file.js",
"exportName": "createResearchAgent",
"factoryArgs": {
"modelProvider": "anthropic",
"modelName": "claude-3-5-haiku-20241022",
"includeMemory": true
}
}
This allows you to import agents from TypeScript/JavaScript files created with the mcp-ai-agent library. The imported file can export:
- A pre-configured agent instance
- A factory function that creates an agent (with optional arguments)
- An async factory function for complex initialization
For custom functionality, you have several options:
1. Use Prebuilt Servers (recommended): Choose from available prebuilt servers like sequentialThinking, memory, braveSearch, fetch, etc.
2. Create Custom MCP Servers: Build your own MCP server and integrate it:
{
"name": "Calculator Agent",
"description": "A calculator agent with custom math operations",
"model": {
"provider": "openai",
"model": "gpt-4o-mini"
},
"toolsConfigs": [
{
"mcpServers": {
"calculator": {
"command": "node",
"args": ["/path/to/your/calculator-mcp-server.js"]
}
}
}
]
}
3. Combine Multiple Prebuilt Servers: Mix and match existing servers for complex functionality:
{
"name": "Multi-Tool Agent",
"description": "An agent with multiple capabilities",
"model": {
"provider": "anthropic",
"model": "claude-3-5-haiku-20241022"
},
"toolsConfigs": [
{
"prebuilt": "sequentialThinking"
},
{
"prebuilt": "memory"
},
{
"prebuilt": "fetch"
}
]
}
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"mcp agent server": {
"mcp-agent-server": {
"command": "node",
"args": [
"/full/path/to/mcp-agent-server/dist/index.js"
]
}
}
}
}McpServers
{
"mcp-agent-server": {
"command": "node",
"args": [
"/full/path/to/mcp-agent-server/dist/index.js"
]
}
}A Model Context Protocol (MCP) hub that seamlessly connects AI clients with intelligent agents, enabling powerful cross-platform communication across multiple AI ecosystems.
Overview
MCP Agent Server creates a bridge between MCP-compatible clients (like Claude Desktop, VS Code, Cursor, etc.) and specialized AI agents. It enables you to:
- Create and configure multiple specialized agents with different capabilities
- Connect these agents to your MCP clients
- Compose agents into master agents for complex workflows
- Build AI applications with advanced capabilities MCP servers
This project leverages the mcp-ai-agent framework to simplify agent creation and management.
Features
- Multiple Agent Support: Create and manage different specialized agents
- Agent Composition: Combine specialized agents into master agents
- Custom Tool Integration: Create your own tools or use existing MCP servers
- Preconfigured Servers: Easy access to popular MCP servers like Sequential Thinking, Brave Search, and Memory
- AI SDK Integration: Support for 17+ LLM providers including OpenAI, Anthropic, Google (Generative AI & Vertex), Amazon Bedrock, Azure, Cohere, Mistral, Fireworks, Groq, Perplexity, Together AI, xAI, DeepSeek, Cerebras, DeepInfra, and Replicate through Vercel AI SDK v5
Prerequisites
- Node.js (v16+) installed
- An API key for your chosen AI model provider (OpenAI, Anthropic, Google, etc.)
- Any API keys required by specific MCP servers you want to use
Installation
Option 1: Using npx (Recommended)
You can run MCP Agent Server directly using npx without installation:
```bash
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
